194 FETCH; 



from the axil of which the shoot arose. In Balansia Bambusss 

 the stroma encloses the inner leaves of the shoot, and produces 

 an external perithecial portion at the apex. In all cases the 

 tissue of the stroma is composite, being formed partly of fungus 

 tissue and partly of the almost unmodified tissues of the host plant, 



Ophiodothis Atkinson, according to Atkinson and von 

 Hohnel, is identical with Balansia Speg. Moller included 

 the species with stalked perithecial heads in Balansia, and the 

 pulvinate forms in Ophiodothis. Dothichloe Atkinson differs 

 only slightly from Balansia, practically only in the thickness 

 of the stroma, and is scarcely worthy of retention. Similarly, 

 Balansiopsis v. Hohnel and Linearistroma v. Hohnel are 

 scarcely distinct from Balansia. If these are retained, it 

 would appear that Moller's distinction between Ophiodothis 

 and Balansia should hold. Ophiodothella v. Hohnel, originally 

 proposed by Hennings as a subgenus of Ophiodothis, appears 

 to belong not to this group, but to the Dothidacese. 



The genera referred to above agree in having a composite 

 stroma, and, with the genus Claviceps, they constitute the 

 Glavicipiteae . They differ from Hypocrella in the structure 

 and colour of the stroma, and in having filamentous septate 

 spores which do not divide into part-spores in the ascus. 

 It appears to be generaUy accepted that the genus Ephelis 

 includes the conidial stages of Balansia, though Moller stated 

 that the conidial stage of Balansia redundans had no resem- 

 blance to Ephelis. Moller obtained a Gephalosporium form on 

 the germination of the ascospores of the latter species, but 

 it does not necessarily follow that that is the form which 

 occurs in nature, or that it is the only conidial form of the 

 species. Immature stromata, which probably belong to this 

 group, are common on the inflorescences of grasses in Ceylon ; 

 they usually consist of white or grayish mycelium, which 

 binds together the individual spikelets into a continuous mass. 



Dussiella Pat. has been attributed in part to Hypocrella, 

 The genus was founded by Patouillard in 1890 on three 

 collections from widely-separated locaUties. The first of 

 these, from North America, had been named Hypocrea 

 tuberifarmis by Berkeley and Ravenel. With this Patouillard 

 joined a specimen from Caracas in Herb. Berlin, and anothei' 



